Background: Schools are responsible for supporting student growth. Various factors impact how students develop at school (Visscher, 2017), but the teacher is most proximal to students. Teachers' decisions inform students' opportunity to learn, engagement, and achievement (Backes et al., 2024; Bergold & Steinmayr, 2023; Chetty et al., 2011; Nye et al., 2004). Thus, it makes sense to invest in improving teachers' practices, often done via teacher professional development (TPD). Researchers have evaluated TPD intervention effects on teaching competence, quality, and student achievement (Kahmann et al., 2022; Garrett et al. 2021). The number of rigorous primary studies assessing TPD's impact has grown, and meta-analyses have increased in number. Like primary studies, however, meta-analyses exhibit significant methodological differences. Meta-analytic findings can play an important role in guiding policymakers' decisions and assist program designers in shaping TPD interventions. It is therefore important to know what has been investigated about TPD across meta-analyses thus far, to determine findings and methods by which those findings were attained. These answers enable reflection on what is known, the degree to which findings are consistent, strengths and weaknesses of the extant research, and needed future directions. Objectives: In this study, we investigated TPD meta-analyses by asking--1. What definitions of TPD have been used in meta-analyses to date? 2. To what extent are TPD effects and their moderators consistent across reviews? Methods: We conducted a review of systematic reviews, focusing on TPD for in-service educators. Eligibility Criteria: To be included, each meta-analysis must have--1. Evaluated the effects of TPD targeting in-service educators in grades PreK-12. 2. Measured effects on student academic achievement. 3. Reported at least average effect size, using a standardized mean difference. 4. Published in English. Search Strategy: To locate studies, we--(1) retrieved previous meta-analyses from a recent review of TPD research (Author et al., 2025; k = 30); (2) updated this list by searching Education Source (through EBSCOhost; k = 71) using the following search string: (meta-analysis OR metanalysis) AND ("teacher professional development" OR "professional learning" OR "teacher development" OR "teacher training" OR "professional development learning"); and (3) manually searched the database of reviews created by Author (2025), which includes all meta-analyses from 2011 to 2023 published on the impact of school-based interventions on academic achievement (k = 247). The PRISMA flowchart (see Figure 1) shows our selection process. After independent double-screening of each study, a total of 14 meta-analyses were included. Data Extraction: We developed a codebook with characteristics grouped in three categories--general findings, TPD characteristics, and methodological features. To mitigate potential sources of bias in the review, we critically examined studies using the following methodological characteristics based on empirically supported review guidelines (Page et al., 2021; Pigott & Polanin, 2020): search strategy, inclusion criteria, critical appraisal, meta-analysis model, and moderator analysis model. The codebook (Appendix A) was piloted by all coders and revised for clarity. Then, three reviewers (two per study) independently extracted data; disagreements were solved by two additional reviewers. We observed considerable variation in moderator terminology. To facilitate meaningful comparisons across studies, we created six categories a posteriori to standardize moderator reporting: (1) method-related characteristics, (2) student sample characteristics, (3) teacher intervention characteristics, (4) outcome characteristics, (5) review-specific characteristics, and (6) other moderators. Results: Table 1 summarizes key methodological characteristics of the included reviews. RQ1: TPD Definitions: Meta-analysts operationalized TPD quite differently across reviews (see Appendix B). Definitions' broadness created a diverse picture of how TPD is defined and studied. Some reviews only referred to improving teacher knowledge, while others also addressed teaching skills. Some reviews addressed multiple skills, while others focused on only one. Some researchers were narrow in their TPD definitions, while some were general. This variation did not enable a straightforward comparison across meta-analyses. RQ2: TPD Effects--On average, TPD interventions positively impacted learner outcomes (see Table 2). Average effect sizes ranged from 0.05 to 0.54 across the meta-analyses. Sixteen of the 18 average effect sizes were positive and statistically significant at a 95% confidence level. Twelve of the 14 reviews modeled heterogeneity. The most frequently studied moderators fell into the "teacher intervention characteristics" category (k =10). Table 3 displays extracted moderators, organized by category. The table shows substantial variability across studies in the moderators examined and their influence. Most of the meta-analyses tested one moderator at a time, which limited the ability to control for the confounding effect of other relevant moderators and covariates. Many intervention characteristics (i.e., length, content, dosage) could not establish statistical significance. However, incorporating teacher collaboration appears to be relevant for achieving better results. This finding aligns with existing research on effective TPD features (e.g., Darling-Hammond et al., 2017; Desimone, 2009). Conclusions: The substantial variability in TPD definitions across studies reflects an absence of a uniform definition within the literature. This lack of consensus influences both study selection and reported outcomes, complicating comparisons across research. The latter also blocks growth in knowledge regarding what makes TPD effective. Given the presence of existing TPD frameworks that provide clear definitions and indicators of high-quality TPD (e.g., Darling-Hammond et al., 2017; Desimone, 2009), TPD meta-analysts should draw upon theoretical frameworks that define TPD clearly. Despite variations in scope and methodologies, the average impact of TPD on student achievement was consistently positive and significant. This provides robust support for TPD investments. It is striking that the TPD moderators studied thus far have been rather superficial, focusing primarily on one aspect of TPD (e.g., length and dosage). The meta-analytical study of effective strategies for improving complex teacher skills could be more profound and better align with what is known about such processes (van Merrienboer & Kirschner, 2017). Researchers should determine the combinations of TPD goals, content, methods, and context that work best for improving teacher practice, and, ultimately, student learning. It is also crucial to apply modern meta-regression methods that allow to model multiple competing moderators to control for confounding (Tipton et al., 2019).