Conventional alcohol catalytic chemical vapor deposition (ACCVD) growth of carbon nanotubes (CNTs) have been carried out under ambient gases from 103 to105 Pa. These ambient gas pressures have prevented in situ observations using an electron beam during CNT growth, such as scanning electron microscopy (SEM), scanning tunneling microscopy (STM). Therefore, in order to realize the in situ observations and to clarify the growth mechanism of nanotube, CNT growth in a high vacuum is essential. In addition, the effects of residual gases also may be avoided in the growth under high vacuum. In this study, we carried out CNT growth under high vacuum using an alcohol gas source in an ultrahigh vacuum (UHV) chamber and we achieved CNT growth below 400°C without any excitation processes of carbon source. After deposition of Co catalyst of 1 nm in thickness on SiO2/Si substrate, ethanol gas was supplied to the substrate surface through a stainless steel nozzle in the UHV chamber. The growth temperature was monitored by a pyrometer during the growth, and set between 350 and 900°C. The supply of ethanol gas was controlled by monitoring an ambient pressure, which was varied from 1 ∼10-1 to 1 ∼10-4 Pa. The grown CNTs were characterized by SEM and Raman spectroscopy. The G/Si intensity ratio reached its maximum at 700°C, when the pressure was 1 ∼10-1 Pa. The maximum point of the G/Si peak intensity shifted to a lower temperature as the growth pressure decreased. When the pressure was 1 ∼10-4 Pa, the G/Si intensity ratio reached its maximum at 400°C, at which clear RBM peaks were observed in the Raman spectrum. From the RBM peaks, the CNT diameters were estimated to be between 0.9 to 1.7 nm, and CNTs of 1.2-1.4 nm in diameter were dominant at 1 ∼10-1 Pa, whereas thinner CNTs (diameter is below 1.0 nm) were increased with the reduction of the pressure. Our largest G/D ratio was about 40 for the sample grown at 1 ∼10-1 Pa, which is considerably larger than the reported value for the CNTs grown under low pressure. From these results, we conclude that the reduction of the growth pressure lowers the growth temperature. This technique can be applied to in situ observation, and may also be useful for low temperature growth of CNTs, which opens new possibilities for the fabrication of CNT based nanodevices.