The gender gap in academic research production has been demonstrated through various methods and geographical contexts. Previous research has yielded conflicting results concerning the effect of gender as a stand-alone variable on research production and primarily focused on other individual and organizational variables to explain this gap. We seek to clarify the influence of gender-related variables on academic research production, operationalized as journal article equivalents published three years prior to the survey. Using a stratified sample of 2,968 full-time Canadian university professors (49% men, 49% women; 30 to 89 years of age, M = 56, SD = 10.17) from the Academic Profession in the Knowledge Society survey, we analyzed the effect of 15 gender-related predictors on research production and measured the variance of research production explained by the combination of these predictors. Through multiple linear regressions, our findings indicate that gender as an isolated variable is a weak yet significant predictor ([beta] = 0.04; p < 0.05) of research production, and that a large proportion of variance (R[superscript 2] = 0.33) is explained by a combination of gender-related variables, such as collaborations, workload division and academic rank. Contrary to what has often been assumed, the gender gap in academic research production is not adequately understood through the effect of gender as a stand-alone variable and gains clarity when we use the 'multilevel approach to gender inequalities' conceptual framework, which enlightens the understanding of this phenomenon as a part of the intersection and reinforcement of individual, interactional, organizational, systemic, and cultural institutional factors.