Abstract Exoplanets with radii between those of Earth and Neptune have stronger surface gravity than Earth, and can retain a sizable hydrogen-dominated atmosphere. In contrast to gas giant planets, we call these planets gas dwarf planets. The James Webb Space Telescope (JWST) will offer unprecedented insight into these planets. Here, we investigate the detectability of ammonia (NH₃, a potential biosignature) in the atmospheres of seven temperate gas dwarf planets using various JWST instruments. We use petitRadTRANS and PandExo to model planet atmospheres and simulate JWST observations under different scenarios by varying cloud conditions, mean molecular weights (MMWs), and NH₃ mixing ratios. A metric is defined to quantify detection significance and provide a ranked list for JWST observations in search of biosignatures in gas dwarf planets. It is very challenging to search for the 10.3–10.8 μm NH₃ feature using eclipse spectroscopy with the Mid-Infrared Instrument (MIRI) in the presence of photon and a systemic noise floor of 12.6 ppm for 10 eclipses. NIRISS, NIRSpec, and MIRI are feasible for transmission spectroscopy to detect NH₃ features from 1.5–6.1 μm under optimal conditions such as a clear atmosphere and low MMWs for a number of gas dwarf planets. We provide examples of retrieval analyses to further support the detection metric that we use. Our study shows that searching for potential biosignatures such as NH₃ is feasible with a reasonable investment of JWST time for gas dwarf planets given optimal atmospheric conditions.