<p>Because of the intensification of production systems and the continued use of concentrated fertilizers without sulfur (S), S deficiency in crops has increased worldwide. We compared the efficiency of S sources applied on an Oxisol in a no-till wheat (<i>Triticum aestivum</i> L.)-soybean [<i>Glycine max</i> (L.) Merrill] cropping system in Southern Brazil. A randomized complete block design was used with four replicates. The treatments were: control (no S), single superphosphate (SSP), elemental S (ES), and phosphogypsum (PG). The amount of S supplied by the sources was 195&#xa0;kg S ha^− 1 in a single application as PG and in three annual applications of 65&#xa0;kg S ha^− 1 for each wheat-soybean succession as SSP and ES. To balance the amount of phosphorus applied with SSP and isolate the effect of S, monoammonium phosphate (MAP) was applied for each wheat-soybean succession at a rate of 100&#xa0;kg P₂O₅ ha^− 1 in the control, ES, and PG treatments. The SO₄-S content in the soil profile efficiently increased with the SSP, ES, and PG applications. The S-leaf content of wheat increased with ES and PG applications, and the leaf-S content of soybean was no affected. Wheat and soybean grain yields were not influenced by SSP, ES, and PG applications, although soybean protein content was increased with the application of S-containing fertilizers. A level of 13&#xa0;mg dm^− 3 of SO₄-S in the soil at 0–0.20&#xa0;m depth was sufficient to supply the S demand in a no-till wheat-soybean cropping system.</p>