[Objectives] Cognition plays an important role in many ecological processes across animals. Growing evidence has indicated individual differences in cognitive performances across multiple species, and the cases of this phenomenon have shown mixed results. However, most of these studies focused on discrete cognitive task or certain limited factors, which may lead to bias. The aims of this study are two-fold: (1) investigate the effects of personality traits (neophobia and exploration), sex, and body characteristics on four general cognitive tasks; (2) investigate whether performances in problem-solving tasks are associated with personality traits, sex, and body characteristics. [Methods] We conducted a series of cognitive experiments to investigate the effects of sex, tarsus length, head volume, and personality traits (exploration and neophobia) on general cognitive tasks and problem-solving performances of budgerigars (Melopsittacus undulatus). Apparatuses with two levels of difficulty were used to measure the problem-solving performances of budgerigars. Exploration behavior was measured in a closed apparatus (L × W × H: 120 cm × 65 cm × 65 cm), which included five artificial trees (height × diameter: 50 cm × 1.5 cm) with two branches (one at the height of 25 cm and one at the top; Fig. 1a). A transparent acrylic sheet was covered on the apparatus top for recording. Budgerigars were caught via flight cages and placed in small cages (L × W × H: 29.3 cm × 22.5 cm × 28 cm), which were coved with black cloth to keep the light out. After 10-min acclimatization, the cage door was opened, and behavior was immediately recorded for 2 min when a budgerigar entered the apparatus. The sum of the walking, flying, and jumping behaviors of each bird in 2 min was regarded as the exploratory score. The test was conducted three times, with the intervals being more than 15 d. In the neophobia test, we recorded the time before individuals began to touch the apparatuses in inhibitory control and associative learning tasks, which were regarded as individuals’ neophobia of different experimental apparatuses. The inhibitory control ability was tested by a detour reaching task. The apparatus consisted of a transparent open-ended cylinder (length × diameter: 7 cm × 7 cm) in which a food reward was placed in the center (Fig. 1b). A lead block was placed at the bottom of the apparatus, which could prevent the bird from easily moving. In each trial, an attempt was considered as successful when the individual detoured the ends of the apparatus without pecking the transparent wall three times. The test was conducted 10 times, with each lasting for 1 min. The number of successful times was seemed as the score in this task. Individuals with high scores were considered to have a better performance in inhibitory control, while a low score indicated poorer performance. The associative learning was tested by a color discrimination task. The apparatus was a rectangular block of wood (L × W × H: 15 cm × 6 cm × 3 cm) with two wells (diameter × depth: 2 cm × 1 cm; Fig. 1c). The wells were covered with dark green and light green lids. In the first trial, an individual needed to open both of the lids and find the food reward, and the reward color was set in light green. The trials were no more than 50 times per day, with the interval of 1 min. If the bird chose the wrong well, the apparatus was immediately removed and the next trial was conducted 1 min later. The individuals were regarded as pass the associative learning task when they found the correct reward well on eight of nine consecutive attempts. The total number of attempts made by the budgerigar was counted as the score for this experiment. A low score indicates better performance in associative learning, while a high score indicates poorer performance. The reversal learning task began 24 h later when the associative learning task was finished. In this task, the food reward was in the dark green well and other experimental procedures were same as those in the associative learning task. The total number of attempts made by the budgerigar was counted as the score for this test. Individuals with low scores are considered to have a better performance in reversal learning, while a high score indicates poorer performance. The spatial memory task was conducted with a rectangular board, which contained 8 wells with blue lids (L × W × H: 20 cm × 15 cm × 3 cm; Fig. 1d). Individuals needed to open the correct lid to get the food reward randomly placed into the well. This test encompassed four phases. In the first phase, individuals had to open the blue lids to search for the reward well. Once they found the correct well and ate the food, the board was removed. After a 5-min interval, the second phase began, with the food reward well being the same as that in the first phase. The first and second phases were training phases that allowed budgerigars to remember the correct reward well. The third and fourth phases were test phases with a 24-h interval, and these two phases were carried out 24 h and 48 h after the second phase, respectively. Budgerigars need to use their memory from the training phases to find the correct rewarded well. If they did not find the reward well within 2 h of the testing phase, we regarded these individuals as non-solvers, and the maximum score in this phase was recorded. The total number of lids opened by budgerigars in the third and fourth phases before they found the correct well was taken as the memory score, and the maximum score was 14. Individuals with low scores were considered to have a better performance in the spatial memory task, and high scores indicate poorer performance. The problem-solving task was carried out via two tests with different apparatuses. The apparatuses were categorized into two types, simple and complex, basing on different levels of difficulty. The simple apparatus was a petri dish with food reward (lid diameter: 70 mm; Fig. 1e), and individuals needed to open the petri dish to obtain the food reward by bill. The complex apparatus was a cuboid made of transparent acrylic sheets with an open end (L × W × H: 15 cm × 6 cm × 6 cm). Inside the complex apparatus was placed a small box with a flat stick that can be pulled out (Fig. 1f). The criterion for the success of this task was that the individuals had to use the stick to pull the box out of the apparatus and then open the lid on the box to obtain the food reward. The simple and complex apparatuses were respectively placed into the cages for 4 days (2 h from 9:00 to 11:00 and 2 h from 15:00 to 17:00 per day), and each apparatus needed to be tested for 16 h. If the individual fails to solve the problem within 16 h, the test will be deemed a failure. The number of attempts, the time before touching the apparatus, and the time to successfully opening the apparatus in problem solving tasks were recorded. In the exploration test, a linear mixed model was employed to calculate the individual repeatability of the exploration. To meet the normal distribution, exploration scores were square-root transformed and used as the response variable. The order (test sequence), sex, time, and date (days since January 1) were used as independent variables. The R package ‘rptR’ was used for data analysis. Then, the Best Linear Unbiased Prediction (BLUP) for each individual was calculated 1 000 times via the R package ‘arm’. In subsequent analyses, the mean BLUP values were used as exploration scores. In the spatial memory task, to demonstrate that budgerigars remembered the location of the reward well, according to previous study, we considered searching more than 4.5 wells as random searching. One-sample t-tests were conducted on the scores of the budgerigars in the third and fourth phases. We adopted six generalized linear mixed models (GLMMs) to analyze the effects of sex, tarsus length, exploration, and head volume on the cognitive performances in the six tests. We added neophobia as an independent variable in the inhibitory control, associative learning, and reversal learning tasks. To investigate the effects of sex and exploration on the problem-solving attempts, we adopted two GLMMs. Furthermore, we added the time before touching the apparatus as an independent variable in the problem-solving (complex apparatus) task. The ID was used as a random effect in all models, and all data analyses were conducted in R version 4.4.2. [Results] The repeatability of exploratory behavior was significant in exploration tests (R = 0.508, P < 0.001), and the confidence interval was [0.333, 0.651]. The average neophobia of individuals towards the inhibitory control apparatus was 373.3 ± 706.1 s, and that towards the associative-reversal learning apparatus was 141.2 ± 412.7 s. In the inhibitory control task, the average number of successful attempts was 7.42 ± 2.35. We found that a smaller head volume corresponded to a higher score in inhibitory control (Fig. 2a; Table 1), and male budgerigars had higher scores than females (Fig. 2b; Table 1). The scores in the inhibitory control task were not significantly related to neophobia or exploration (Table 1). In the associative learning task, the average number of attempts to meet the criterion was 46.84 ± 31.10. Individuals with smaller head volumes had lower number of attempts (Fig. 2c; Table 1), and male budgerigars had lower scores than females (Fig. 2d; Table 1). The attempts were not significantly related to neophobia, sex, exploration, or tarsus length (Table 1). In the reversal learning task, the average number of attempts to meet the criterion was 60.56 ± 24.07. The attempts were not significantly related to neophobia, sex, exploration, head volume, or tarsus length (Appendix 1). In the spatial memory task, the average number of attempts to find the reward well was 4.04 ± 3.39. No significant associations were found between measured factors and spatial memory scores (Appendix 1). In the problem-solving task (simple apparatus), the problem was solved by problem solvers in 3.58 ± 3.77 h (n = 40), and the non-problem solvers did not solve the problem within 16 h (n = 12). The number of attempts was 15.58 ± 11.54 (n = 52). Among the birds, fast explorers were more likely to open the apparatus (Fig. 3a; Table 2). Females were more likely to open the apparatus than males (Fig. 3b; Table 2). Budgerigars with longer tarsus were more likely to open the apparatus (Table 2). Whether the budgerigar could open the apparatus was not significantly related to the head volume (Table 2). Among the 40 budgerigars that successfully solved the problem, females spent shorter time to open the apparatus (Fig. 3b; Table 2), and the time was not significantly related to exploration, tarsus length, or head volume (Table 2). The number of attempts was not significantly correlated with sex among all budgerigars or only problem solvers (Appendix 2). In the problem-solving task (complex apparatus), the problem was solved by problem solvers in 3.49 ± 3.22 h (n = 30), and the non-problem solvers did not solve the problem within 16 h (n = 22). The number of attempts was 20.35 ± 16.54 (n = 52), and the time before touching the apparatus was 769.8 ± 1 890.4 s. Among the birds, females were more likely to solve the problem than males (Fig. 3c; Table 3). Whether the budgerigar could open the apparatus was not significantly related to exploration, tarsus length, the time before touching the apparatus, or head volume (Table 3). Among the 30 budgerigars that successfully solved the problem, individuals with longer tarsus spent shorter time to open the apparatus, and the time was not significantly related to sex, exploration, the time before touching the apparatus, or head volume (Table 3). Among the budgerigars that successfully solved the problem, females had a higher number of attempts than males (estimate = 14.410, SE = 5.692, df = 26, χ2 = 6.409, P = 0.011; Fig. 3d), while the number of attempts was not significantly correlated with exploratory behavior or the time before touching the apparatus (Appendix 3). Among all budgerigars, the number of attempts did not significantly correlate with exploration, sex, or the time before touching the apparatus (Appendix 3). [Conclusion] The findings indicate that cognition is related to personality, sex, and physical conditions, which enrich the understanding of cognition and its influencing factors. In the future, more in-depth experiments should be conducted, such as additional personality tests to examine their interrelationships and connections to cognitive performance.