Background: The painDETECT questionnaire (PD-Q) is a fast and uncomplicated way to ascertain the percentage of neuropathic pain in ‘total pain’ and is designed to detect neuropathic pain components in back pain. The purpose of this randomized, prospective study is to compare, with the assessment of the PD-Q, the efficacy of interlaminar (IL) and transforaminal (TF) steroid injections in patients with unilateral chronic lumbar radicular pain. Methods: Patients were treated fluoroscopically with epidural steroids, using the IL or TF method and with confirmation of the epidural space by contrast, using random computerized classification. The patients received a series of three IL or TF epidural steroid injections (ESI) at 2-week intervals. The patients were monitored for 6 months from the first steroid injection. Results: By analyzing the average values of the total sum of points in the PD-Q a dropping trend is confirmed for both groups. The trend equation (y = –1.1393x + 25.269) for the TF ESI shows a faster recovery than the IL ESI (y = –0.8089x + 26.654). The statistically significant difference in the two groups is proved between the first and the sixth visit (IL ESI, p = 0.014; TF ESI, p = 0.001). There is no statistically significant difference in the efficiency of the two dosages and the volumes of steroids between the IL and TF distribution of steroids. Conclusions: Steroids are efficient; besides alleviating the overall pain, they also reduce the neuropathic component in chronic lumbar radicular pain, whether it is distributed epidurally by the IL or TF approach.