The value of radiotherapy to the chest (RC) in disseminated small cell lung carcinoma (SCLC) has been questioned. Two protocols for disseminated SCLC from the Southwest Oncology Group (SWOG) have been compared. They were developed four years apart. The first one included radiotherapy (RT), 3000 rad in two weeks, to the primary tumor, mediastinum and supraclavicular areas, while the second one deleted any RC. Multidrug chemotherapy (CT) and brain RT were used in both protocols. Nonresponders to CT were removed from the study. Our main findings are as follows: (1) Initial chest relapses (patients with no initial extrathoracic relapse) have increased from 24–55% when RC is not given (P = 0.0001). Overall chest relapse (adding those patients that relapsed simultaneously in the chest plus other sites) in the second protocol was 73%. (2) Amount of response to CT does not influence the chances for relapse. Even complete responders to CT have a high chance for chest relapse. (3) Sites of relapse without RC are mainly in the primary tumor, ipsilateral hilus and mediastinum. (4) With RC, relapses shift to the chest periphery, mostly to the lung outside the radiotherapy field and to the pleura. (5) The two very different CT regimens have produced similar percentages and duration of response. (6) CT schema with periodic reinductions prolongs duration of response and survival over schema with continuous maintenance. Hence, interruption of CT to allow RC does not seem to adversely influence CT efficacy. From our results and the review of the literature we conclude that: (1) patients with disseminated SCLC that respond to CT should be given RC to decrease chest relapses. (2) A dose of 3000 rad in two weeks seems to be enough to produce a low percentage of chest relapse in disseminated SCLC, as survival of these patients is short and many will die prior to developing chest relapse. However, according to the literature, 4000–4800 rad is probably a more effective dose. (3) More studies and guidelines are needed to outline proper boundaries of radiotherapy fields, to decrease chances of peripheral chest relapses. (4) Median survival might not be a good parameter to evaluate the impact of RC in disseminated SCLC. The study of long‐term survivors seems to be more important.