The new TomoDirect (TD) modality offers a nonrotational option with discrete beam angles. We aim to compare dosimetric parameters of TD, helical tomotherapy (HT), volumetric-modulated arc therapy (VMAT), and fixed-field intensity-modulated radiotherapy (ff-IMRT) for upper thoracic esophageal carcinoma (EC). Twenty patients with cT2-4N0-1M0 upper thoracic esophageal squamous cell carcinoma (ESCC) were enrolled. Four plans were generated using the same dose objectives for each patient: TD, HT, VMAT with a single arc, and ff-IMRT with 5 fields (5F). The prescribed doses were used to deliver 50.4 Gy/28F to the planning target volume (PTV50.4) and then provided a 9 Gy/5F boost to PTV59.4. Dose-volume histogram (DVH) statistics, dose uniformity, and dose homogeneity were analyzed to compare treatment plans. For PTV59.4, the D2, D98, Dmean, and V100% values in HT were significantly lower than other plans (all p < 0.05), and those in TD were significantly lower than VMAT and ff-IMRT (all p < 0.05). However, there was no significant difference in the D2 and Dmean values between VMAT and ff-IMRT techniques (p > 0.05). The homogeneity index (HI) differed significantly for the 4 techniques of TD, HT, VMAT, and ff-IMRT (0.03 ± 0.01, 0.02 ± 0.01, 0.06 ± 0.02, and 0.05 ± 0.01, respectively; p < 0.001). The HI for TD was similar to HT (p = 0.166), and had statistically significant improvement compared to VMAT (p < 0.001) and ff-IMRT (p = 0.003). In comparison with the 4 conformity indices (CIs), there was no significant difference (p > 0.05). For PTV50.4, the D2 and Dmean values in HT were significantly lower than other plans (all p < 0.05), and those in TD were significantly lower than VMAT and ff-IMRT (all p < 0.05). However, there was no significant difference in the D2 and Dmean values between VMAT and ff-IMRT techniques (p > 0.05). No D98 and V100% parameters differed significantly among the 4 treatment types (p > 0.05). HT plans were provided for statistically significant improvement in HI (0.03 ± 0.01) compared to TD plans (0.05 ± 0.01, p = 0.003), VMAT (0.08 ± 0.03, p < 0.001), ff-IMRT (0.08 ± 0.01, p < 0.001). The HI revealed that TD was superior to VMAT and ff-IMRT (p < 0.05). The CI differed significantly for the 4 techniques of TD, HT, VMAT, and ff-IMRT (0.59 ± 0.10, 0.69 ± 0.11, 0.64 ± 0.09, and 0.64 ± 0.11, respectively; p = 0.035). The best CI was yielded by HT. We found no significant difference for the V5, V10, V15, V30, and the mean lung dose (MLD) among the 4 techniques (all p > 0.05). However, the V20 differed significantly among TD, HT, VMAT, and ff-IMRT (21.50 ± 7.20%, 19.50 ± 5.55%, 17.65 ± 5.45%, and 16.35 ± 5.70%, respectively; p = 0.047). Average V20 for the lungs was significantly improved by the TD plans compared to VMAT (p = 0.047), and ff-IMRT (p = 0.008). The V5 value of the lung in TD was 49.30 ± 13.01%, lower than other plans, but there was no significant difference (p > 0.05). The D1 of the spinal cord showed no significant difference among the 4 techniques (p = 0.056). All techniques are able to provide a homogeneous and highly conformal dose distribution. The TD technique is a good option for treating upper thoracic EC involvement. It could achieve optimal low dose to the lungs and spinal cord with acceptable PTV coverage. HT is a good option as it could achieve quality dose conformality and uniformity, while TD generated superior conformality.