Abstract Soil acidification induced by reactive nitrogen (N) inputs can alter the structure and function of terrestrial ecosystems. Because different N‐transformation processes contribute to the production and consumption of H + , the magnitude of acidification likely depends on the relative amounts of organic N (ON) and inorganic N (IN) inputs. However, few studies have explicitly measured the effects of N composition on soil acidification. In this study, we first conducted a meta‐analysis to test the effects of ON or IN inputs on soil acidification across 53 studies in grasslands. We then compared soil acidification across five different ON:IN ratios and two input rates based on long‐term field N addition experiments. The meta‐analysis showed that ON had weaker effects on soil acidification than IN when the N addition rate was above 20 g N m −2 year −1 . The field experiment confirmed the findings from meta‐analysis: N addition with proportions of ON ≥ 20% caused less soil acidification, especially at a high input rate (30 g N m −2 year −1 ). Structural equation model analysis showed that this result was largely due to a relatively low rate of H + production from ON as NH 3 volatilization and uptake of ON and NH 4 + by the dominant grass species Leymus chinensis (which are both lower net contributors to H + production) result in less NH 4 + available for nitrification (which is a higher net contributor to H + production). These results indicate that the evaluation of soil acidification induced by N inputs should consider N forms and manipulations of relative composition of N inputs may provide an effective approach to alleviate the N‐induced soil acidification.