The MBTI is not well regarded by actual psychologists. The National Academy of Sciences has ruled that "At this time there is not sufficient, well-designed research to justify the use of the Myers-Briggs Type Indicator (MBTI) in career counseling programs." (p15). http://www.nap.edu/openbook.php?record_id=1580
"Myers and McCaulley (1985) report test-retest reliabilities from 11 different samples. In these samples, between 24 and 61 percent of the respondents showed stability of type, with a median of 40 percent... McCarley and Carskadon (1983) report only 47 percent of their respondents retained their initial types over a period of 5 weeks... Self-assessment instruments like those represented by the MBTI are likely, at best, to capture a person 's current state and, therefore, should not be considered typologies." (p96-97)
"Myers and McCaulley (1985) present correlational data from 20 studies showing relationships between MBTI scales and a wide variety of other instruments. They concluded that the I-E scale has good construct validity: it has high correlations with comparable scales of other instruments and low correlations with instruments designed to represent other constructs. In contrast, the S-N and T-F scales show relatively weak validity, with only moderate correlations with other instruments designed to measure similar constructs." (p98)
"[Thayer] notes three methodological criticisms of the research: the reporting of data is inconsistent and incomplete; statistical analyses are often incompletely described and may violate convention (e.g., no overall test of significance is calculated before detailed comparisons are made and no attention is given to appropriate baseline data in the analyses); and subjects or judges are often made aware of the hypotheses being tested. These problems prevent drawing conclusions from the studies, many of which are published in the Journals of the Center for the Applications of Psychological Type, Inc. (Journal of Psychological Type and Research in Psychological Type)*; we also note that, to date, few studies appear in mainstream research journals." (p88)
*This is the company that distributes the MBTI and sells training programs for it.
And, to further the comparison to horoscopes:
"What accounts for the popularity of an instrument that is not yet supported by research? Several reasons can be suggested. From the respondents' perspective, the MBTI provides generally positive feedback: the descriptors do not have clinical implications, they are presented as unique positive attributes, and they are sufficiently vague to apply to a large number of people in a wide variety of situations." (p100)
If you need a cute test to massage your ego, go take some "Which Kardashian Are You?" quizzes on Buzzfeed instead of pretending that you're doing Very Rational And Scientific Psychology. And Jung is cool, but what he does is not really science.