Stage 2 Registered Report: Stress regulation via being in nature and social support in adults, a meta-analysis **Summary** We conducted a **Registered Report** meta-analysis to provide our best estimate of the effect size of two stress regulation strategies: **Being in nature** and **Emotional social support**. To study the effect of these strategies on stress, we studied the affective (characterized by feelings of nervousness, strain, and tensions), physiological (characterized by the activation of the hypothalamic–pituitary–adrenal axis), and cognitive (characterized by rumination and perseverative thinking) components of the stress response. Because stress can have long-term consequences if not kept under control, we also included an assessment of the affective consequences of stress (such as depression and chronic anxiety). [Study Rationale + Hypotheses][2] for details). To ensure an exhaustive coverage of the stress literature, we searched the literature using three databases: ProQuest (an online platform that integrates the results coming from three databases: APA PsycArticles, APA Psycinfo, ProQuest Dissertations & Theses Global‎), PubMED, and Scopus. For every database, we used a combination of keywords to find as many experimental and observational studies as possible, all of which highlighted a link between the two strategies and different components of stress (physiological, affective, and cognitive) and longer-term affective consequences of stress. All studies included human participants (see the [Methods][3] for precise information). We addressed publication bias with an innovative combination of techniques to adjust for the impact of publication bias as well as possible: First, the **p-curve** method estimated whether the studies included in our meta-analysis showed evidential value. Second, the **multiple-parameter selection model** (McShane et al., 2016) and third, the mixed effect implementation of **PET-PEESE** (Stanley & Doucouliagos, 2014) provided bias-corrected ES of the effect of interest. To handle dependencies among the effects, the bias correction methods were implemented using a permutation-based procedure or using multilevel modeling. Finally, we used a robust **Bayesian model-averaging** approach to integrate the selection modeling and regression-based approaches and let the data determine the contribution of each model by its relative predictive accuracy to fit the observed data (Bartoš et al., 2021).Regarding the inference about our target effects, the 3- or 4-parameter selection model (4PSM; McShane, Böckenholt, & Hansen, 2016) was used as the primary inferential and estimation bias-adjustment method (see the [Data Analytic Plan (Before Registration)][4] for precise information). In [Materials][5] you can find all the documentation that were used in the present meta-analysis. *Author Note: The preparation of this work was partly funded by a French National Research Agency ”Investissements d’avenir” program grant (ANR-15-IDEX-02) awarded to Hans IJzerman, and PRIMUS/20/HUM/009 grant awarded to Ivan Ropovik. Note that we wrote the pre-registration in past tense to avoid errors after finishing the manuscript.* [2]: https://osf.io/5bej6/ [3]: https://osf.io/cgy4h/ [4]: https://osf.io/vskpb/ [5]: https://osf.io/4cjux/