Background: Soft tissue sarcomas (STS) are uncommon malignancies with significant biological and clinical variation. After primary curative therapy, patients enter a program of surveillance for recurrence with periodic chest x-rays (CXR) or computed tomography (CT). We compared costs and health outcomes of CXR and CT at a range of frequencies and durations of follow-up. Methods: We used a Markov model to simulate a cohort of 10,000 STS patients through their surveillance experience for lung metastasis after completion of definitive treatment for stage II or III primary disease. Health states in the model were no evidence of disease, recurrence, and death. We assessed the method of chest imaging, duration of follow-up, and interval (every 3 months or 6 months) of surveillance in the first 3 years. Cost effectiveness was assessed for each screening modality and screening frequency. Recurrence probabilities, utilities, treatment costs, and other parameters were from previously published data. Outcomes were costs, quality-adjusted life-years (QALYs) and incremental cost-effectiveness ratios (ICER). Results: The initial evaluation comparing screening every 3 months for 3 years, every 6 months for years 4 and 5, and annually from years 6 to 10 resulted in 632,264 QALYs and $1,038,351,481 costs for CT, compared to 631,834 QALYs and $746,019,937 for CXR, resulting in an ICER of $679,322/QALY. In comparing screening intervals, less frequent screening intervals of every 6 months compared to every 3 months for the first three years using CT resulted in an ICER of $690,527/QALY and for CXR, the ICER was $271,423/QALY. In comparing screening duration between 6 years and 10 years of follow-up, strategies with longer follow-up resulted in slightly higher QALYs in each of the comparison scenarios, at a much higher cost. Conclusions: In our evaluation, more frequent screening the first 3 years and longer duration of surveillance resulted in higher QALYs in both screening modalities. However, in the comparisons the ICERs exceed common willingness to pay thresholds of $150,000/QALY gained; CXR is the more cost-effective imaging modality. Limitation of this model includes the simplification of disease progression and heterogeneity in STS.