ABSTRACT We present the weak-lensing analysis of 279 CODEX clusters using imaging data from 4200 deg2 of the DECam Legacy Survey (DECaLS) Data Release 3. The cluster sample results from a joint selection in X-ray, optical richness in the range 20 ≤ λ < 110, and redshift in the range 0.1 ≤ z ≤ 0.2. We model the cluster mass (M200c) and the richness relation with the expression $\left〈 M_{\rm 200c} | λ \right〉 ∝ M_{0} \, (λ / 40)^{F_{λ }}$. By measuring the CODEX cluster sample as an individual cluster, we obtain the best-fitting values, $M_{0} = 3.24^{+0.29}_{-0.27} \times 10^{14} \text{M}_{⊙ }$, and $F_{λ } = 1.00 ^{+0.22}_{-0.22}$ for the richness scaling index, consistent with a power-law relation. Moreover, we separate the cluster sample into three richness groups; λ = 20–30, 30–50, and 50–110, and measure the stacked excess surface mass density profile in each group. The results show that both methods are consistent. In addition, we find an excellent agreement between our weak lensing based scaling relation and the relation obtained with dynamical masses estimated from cluster member velocity dispersions measured by the SDSS-IV/SPIDERS team. This suggests that the cluster dynamical equilibrium assumption involved in the dynamical mass estimates is statistically robust for a large sample of clusters.