Aims: In order to assess the long term safety, clinical performance and the bioabsorption process of the paclitaxel-eluting absorbable magnesium Scaffold (DREAMS) three-year clinical data and multi-modality imaging outcomes are reported. Methods and results: Forty-six subjects were enrolled in the first-in-man BIOSOLVE-I study in two different cohorts with clinical follow-up at 1, 6, 12, 24 and 36 months; angiographic and IVUS follow-up for cohort 1 at 6-month and for cohort 2 at 12-month. A subgroup of patients underwent OCT and vasomotion testing. The primary endpoint is Target Lesion Failure (TLF) at 6-month for cohort 1 and at 12-month for cohort 2. For some patients also 18-month and 24-month imaging data are available. TLF rate at 36-month was 6.8% including 2 TLRs and 1 peri-procedural MI occurring at the 12-month follow-up angiography; no events emerged from 12- to 36-month. No cardiac death or scaffold thrombosis was observed. Vasoconstriction after acetylcholine at 6-month (delta=-10.04%; p=0.0008 versus baseline) followed by vasodilatation after nitroglycerine (delta=8.69%; p<0.0001 versus baseline) demonstrates the uncaging aspect of the absorption process with no further change at the 12-month follow-up. Six-month virtual histology (VH) data showed a significant decrease in the dense calcium by 39.5% (p=0.0015) remaining stable from 6- to 12-month follow-up. This decrease is interpreted as a surrogate assessment for the bioabsorption process of the scaffold material. Echogenicity data using the decrease in intensity of the ultrasound signal to quantify the change in strut structure demonstrate a continuous decrease in % hyperechogenicity over the follow-up period, with the most pronounced changes within the first 6 months (22 to 16% p<0.001). Conclusions: DREAMS shows excellent safety and efficacy data with no death and no scaffold thrombosis up to 3 years in the BIOSOLVE-I trial. Multi-modality imaging documented the absorption process and the uncaging aspect of this device already at 6 months.