Does distant healing/prayer work?

The evidence picture is more nuanced than a simple pro/con split. The two largest individual clinical RCTs — the STEP prayer trial (Benson 2006, N=1,802) and MANTRA II (Krucoff 2005) — found no benefit of intercessory prayer, and the Cochrane systematic review (Roberts et al. 2009, 10 RCTs, N=7,646) found no clear effect on any primary outcome. However, the most comprehensive meta-analyses tell a different story: Roe et al. (2015) found significant positive effects across both non-whole-human biological systems (r = .115 quality-filtered, 22 studies) and whole-human clinical trials (r = .224 quality-filtered, 27 studies), even after excluding the discredited Wirth studies. These meta-analytic effect sizes are small but robust to quality filtering, though publication bias and quality-outcome correlations temper confidence. The laboratory DMILS literature adds further support: Schmidt's meta-analyses (2004, 2012) find small but significant effects, and Achterberg et al. (2005) found significant fMRI brain activation in recipients during DI intervals. Prayer studies specifically show the weakest effects (r = .138 cropped), while therapeutic touch and Reiki/Johrei show somewhat larger effects. The critical question is whether the positive meta-analytic and laboratory findings reflect genuine healing effects that large individual prayer RCTs are insensitive to, or whether the positive aggregate results reflect publication bias and the methodological flexibility of smaller studies.